This project is an attempt to study changes in patterns of psychoactive drug use by students over time, to derive empirically a classification of user types defined by their patterns of drug use, and to construct a causal model which specifies the process of becoming particular types of drug users and specifies the determinants of changes in usage patterns. It will attempt to do this by means of longitudinal and cross sectional survey and personality data obtained from Carnegie- Mellon University students. The model produced would permit quantitative assessment of the proportion of the total variance inherent in a given pattern of drug use which is attributable to the several most salient independent variables. By use of causal analysis techniques an attempt will be made to construct causal models of drug use which would be empirically quantifiable, verifiable and verified; would reflect a dynamic rather than a static representation of drug use patterns; and would readily lend themselves to the making of predictions in simulation studies.